1. Shape-controlled, high fill-factor microlens
arrays fabricated by a 3D diffuser lithography
and plastic replication method
Sung-Il Chang and Jun-Bo Yoon
3D Micro-Nano Structures Laboratory, Department of Electrical Engineering and Computer Science
Korea Advanced Institute of Science and Technology (KAIST)
373-1 Guseong-dong, Yuseong-gu, Daejeon 305-701, Korea
sichang@3dmems.kaist.ac.kr
Abstract: This paper describes a simple and effective method to fabricate a
plastic microlens array with controllable shape and high fill-factor, which
utilizes the conventional lithography and plastic replication. The only
difference from conventional lithography is the insertion of a diffuser that
randomizes paths of the incident ultraviolet (UV) light to form lens-like 3D
latent image in a thick positive photoresist. After replication of the
developed concave microlens mold onto the polydimethylsiloxane (PDMS),
the focal length of the fabricated hemispherical microlens was observed to
be 13-88 µm depending on the UV exposure dose. Two PDMS curing
conditions were tested, where the elevated temperature of 85 °C resulted in
smoother surface roughness of 2.6 nm in RMS value in the microlens mold.
The proposed method can be extensively applied for microlens fabrication
with other plastic materials due to its simplicity and versatility.
2004 Optical Society of America
OCIS codes: (220) Optical design and fabrication. (220.4000) Microstructure fabrication.
(230) Optical devices. (230.3990) Microstructure devices.
References and links
1. D. Daly, “Applications and Fabrication Technologies” in Microlens Array (Talyer and Francis, London,
2001).
2. M. V. Kunnavakkam, F. M. Houlihan, M. Schlax, J. A. Liddle, P. Kolodner, O. Nalamasu, and J. A.
Rodgers, “Low-cost, low-loss microlens arrays fabricated by soft-lithography replication process,” Appl.
Phys. Lett. 82, 1152-1154 (2003).
3. C. R. King, L. Y. Lin, and M. C. Wu, “Out-of-Plane Refractive Microlens Fabricated by Surface
Micromachining,” IEEE Photonics Technol. Lett. 8, 1349-1351 (1996).
4. M.-H. Wu and G. M. Whitesides, “Fabrication of two-dimensional arrays of microlenses and their
applications in photolithography,” J. Micromech. Microeng. 12, 747-758 (2002).
5. M.-K. Wei and I -L. Su, “Method to evaluate the enhancement of luminance efficiency in planar OLED
light emitting devices for microlens array,” Opt. Express 12, 5777-5782 (2004).
6. J. Chen, W. Wang, J. Fang, and K. Varahramyan, “Variable-focusing microlens with microfluidic chip,” J.
Micromech. Microeng. 14, 675-680 (2004).
7. H. Yang, C.-K. Chao, M.-K. Wei and C.-P. Lin, “High fill-factor microlens array mold insert fabrication
using a thermal reflow process,” J. Micromech. Microeng. 14, 1197-1204 (2004).
8. Z. L. Liau, V. Diaduik, J. N. Walpole and D. E. Mull, “Gallium phosphide microlenses by mass transport,”
Appl. Phys. Lett. 55, 97-99 (1989).
9. Z. L. Liau, D. E. Mull, C. L. Dennis, R. C. Williamson and R. G. Waarts, “Large-numerical-aperture
microlens fabrication by one-step etching and mass-transport smoothing,” Appl. Phys. Lett. 64, 1484-1486
(1994).
10. A. W. Lohmann and S. Singzinger, “Array of doublet lenslets,” Optik 93, 84-86 (1993).
11. R. Yang and W. Wang, “Out-of-plane polymer refractive microlens fabricated based on direct lithography
of SU-8,” Sensor. Actuat. A-Phys. 113, 71-77 (2004).
12. N. Chronis, G. L. Liu, K.-H. Jeong, and L. P. Lee, “Tunable liquid-filled microlens array integrated with
microfluidic network,” Opt. Express 11, 2370-2378 (2003).
(C) 2004 OSA 13 December 2004 / Vol. 12, No. 25 / OPTICS EXPRESS 6366
#5799 - $15.00 US Received 23 November 2004; revised 4 December 2004; accepted 6 December 2004
2. 13. J.-B. Yoon, B.-I. Kim, Y.-S. Choi, and E. Yoon, “3-D Construction of Monolithic Passive Components for
RF and Microwave ICs Using Thick-Metal Surface Micromachining Technology,” IEEE T. Microw.
Theory 51, 279-288 (2003).
14. B. D. Gates and G. M. Whitesides, “Replication of Vertical Features Smaller than 2 nm by Soft
Lithography,” J. Am. Chem. Soc. 125, 14986-14987 (2003).
15. P. Nussbaum, R. Völkel, H. P. Herszig, M. Eisner, and S. Haselbeck, “Design, fabrication and testing of
microlens arrays for sensors and microsystems,” Pure Appl. Opt. 6, 617-636 (1997)
1. Introduction
Microlens arrays have a variety of applications in optical interconnection and optical
communication [1]. Accordingly, various microlens array fabrication methods have been
extensively investigated, including photoresist thermal reflow method [2-7], mass transport
technique [8,9], droplet method [10], and direct lithography method [11]. Also, microlens
arrays with variable focal lengths were developed by liquid-filling in a cavity [6,12]. Among
them, the photoresist reflow method has been most widely used. Nevertheless, due to the
thermal and chemical instability of the photoresist, it is not yet accepted as a practical solution
suitable for commercial optical systems. And reflow conditions must be delicately controlled
to create a high fill-factor microlens array [7], which increases the process uniformity and
reproducibility requirements. Finally, the microlens material should be photosensitive to be
patterned. Other microlens fabrication methods mentioned above had the relatively complex
process steps and were restricted to narrow application range.
From this perspective, 3D diffuser lithography (simply adding a diffuser into
conventional lithography) followed by plastic replication is proposed here to provide a simple
method of fabricating versatile microlens arrays with high fill-factor and pre-defined shape.
2. Microlens fabrication
Figure 1 shows the proposed microlens fabrication processes. Contrary to conventional
lithography in which only a collimated light source is used to obtain photoresist patterns with
a rectangular cross-section, the proposed 3D diffuser lithography utilizes randomized light to
form various cross-sections (circular or elliptical) in the photoresist. After thick AZ9260
photoresist spin-coating (1500 rpm, 1.5 sec for 45 µm thickness) on the substrate and soft-
baking (85 °C, 1 hour), the thick photoresist film was ultraviolet (UV)-exposed through a
diffuser and a mask (Fig. 1(a)). Since the diffuser on the mask randomized the direction of the
UV light, the exposed region formed a circular or an elliptical cross-section after development
(Fig. 1(b)). A liquid polydimethylsiloxane (PDMS) was cast on the photoresist pattern (Fig.
1(c)) and peeled off from the photoresist mold after it was cured (Fig. 1(d)). The curing
condition is described later.
Various cross-sectional photoresist profiles were obtained by using different diffusers.
The type-I diffuser (F43-719, Edmund Optics Ltd.) is a 10 cm × 10 cm square soda-lime 5
mm-thick glass-sheet, of which a 450 µm-thick opaque opal layer is coated on the single-side.
The type-II diffuser (F45-656, Edmund Optics Ltd.) is a 1.6 mm-thick single-side sandblasted
sheet (peak-to-peak roughness=8.7 µm) of the same size. In all experiments, the diffuser was
placed with the opal-coated or sandblasted side up, as indicated in Fig. 1(a). The photoresist
pattern cross-section obtained using the type-I diffuser is shown in Fig. 2(a). The exposure
dose was 300 mJ/cm2
and it had a circular cross-section. The cross-sectional profiles of the
photoresist patterns are overlapped in Fig. 2(b). The photoresist pattern without a diffuser had
a very steep profile on the pattern edges [13], while those with the type-I and II diffusers had
rounded cross-sections.
The exposure dose is another important parameter affecting cross-sectional photoresist
profiles. Figure 2(c) shows the cross-sectional photoresist profiles obtained using the type-I
diffuser (the same diffuser as Fig. 2(a)) and an exposure dose of 1200 mJ/cm2
. The photoresist
pattern cross-sectional profiles shown in Fig. 2(d) indicate that the depth of the photoresist
(C) 2004 OSA 13 December 2004 / Vol. 12, No. 25 / OPTICS EXPRESS 6367
#5799 - $15.00 US Received 23 November 2004; revised 4 December 2004; accepted 6 December 2004
3. patterns increased in proportion to the exposure dose. The maximum incident angle of the UV
light into the photoresist was maintained at about 36° regardless of the exposure dose, as
shown in Fig. 2(d). This is in accordance with the calculation result based on the well-known
Snell’s law when the total internal reflection condition of the light and the refractive indices of
the glass diffuser (ndiffuser=1.53), mask (nmask=1.53), and the photoresist (nphotoresist =1.67) were
considered.
Fig. 1. Process flow of the proposed plastic microlens fabrication. (a) UV exposure through a diffuser
and a mask. (b) Development. (c) Liquid PDMS casting and curing. (d) Replicated PDMS microlens
peel-off.
Fig. 2. SEM photographs and schematic cross-sections of the photoresist patterns. (a) Photoresist
pattern cross-section with type-I diffuser and an exposure dose of 300 mJ/cm2
. (b) Cross-sectional
photoresist profiles (solid line: type-I diffuser, dotted line: type-II diffuser, dashed line: no diffuser). (c)
Photoresist pattern cross-section with type-I diffuser and an exposure dose of 1200 mJ/cm2
. (d) Cross-
sectional photoresist profiles with different exposure doses.
3. Results and discussion
Figure 3 shows SEM photographs of the photoresist patterns with various cross-sections
obtained by an appropriate combination of the pattern width and the exposure dose. Various
diameters of circular cross-sections from 1.5 µm to 80 µm for this experiment were obtained.
All the circular cross-sections in Figs. 3(a) and 3(b) were formed with the type-I diffuser. The
cross-section in Fig. 3(c) was obtained with a wide mask pattern (30 µm), low exposure dose
(300 mJ/cm2
), and the type-I diffuser, while the elliptical cross-section shown in Fig. 3(d) was
(C) 2004 OSA 13 December 2004 / Vol. 12, No. 25 / OPTICS EXPRESS 6368
#5799 - $15.00 US Received 23 November 2004; revised 4 December 2004; accepted 6 December 2004
4. formed by applying a narrow mask pattern (10 µm), high exposure dose (1200 mJ/cm2
), and
the type-II diffuser. Furthermore, we could obtain photoresist patterns with a high density by
using type-II diffuser and quite large exposure dose of 1800 mJ/cm2
. Due to the excess
exposure dose, the adjacent patterns were merged together and such patterns could act as an
excellent photoresist mold for the microlens with a fill-factor of about 100%.
Fig. 3. SEM photographs of the photoresist pattern cross-sections. (a) A circular cross-section with
a 1.5 µm diameter and (b) with an 80 µm diameter. (c) A concave cross-section. (d) An elliptical
cross-section. (e) A cross-section with a very high density.
Fig. 4. SEM photographs of the fabricated plastic microlens array. (a) Conventional convex microlens.
(b) Hemispherical microlens. (c) Ellipsoidal microlens. (d) Microlens array with a high fill-factor
(~100 %).
Based upon extensive photoresist patterning experiments, diverse plastic microlens arrays
were fabricated and their SEM photographs are shown in Fig. 4. The fabricated convex,
hemispherical, and ellipsoidal microlens arrays can be used in appropriate applications.
Especially, the hemispherical microlens array shown in Fig. 4(b) may have a very large
(C) 2004 OSA 13 December 2004 / Vol. 12, No. 25 / OPTICS EXPRESS 6369
#5799 - $15.00 US Received 23 November 2004; revised 4 December 2004; accepted 6 December 2004
5. numerical aperture. The focal length of the fabricated hemispherical microlens was observed
from 88 µm to 13 µm as the exposure dose was increased from 150 to 600 mJ/cm2
. It was
obtained by the radius of curvature measured from the SEM photograph of the individual
microlens, using the conventional formula for a focal length (focal length = radius of
curvature/(refractive index of the lens – 1)).
The microlens array with an extremely high fill-factor was also realized as shown in Fig.
4(d). The fill-factor of the microlens array was almost 100%. And the light energy passing
through the microlens array is expected to be collected with high efficiency if scattering of the
incident light at the sharp borders between two adjacent microlenses is negligible. Thus, the
efficiency of optical systems can be improved.
Table I. Characteristic comparison of the fabricated PDMS microlenses and photoresist molds cured at
two different conditions.
Curing temperature Room temperature 85 °C
Curing time 24 hours 1 hour
SEM photograph of the
fabricated PDMS
microlens
SEM photograph of the
photoresist mold
after PDMS peel-off
AFM measurement
result of the photoresist
mold
after PDMS peel-off
PDMS shrinkage
after peel-off
None 2.2 % shrinkage
During microlens fabrication, the PDMS was cured at room temperature. However, the
curing temperature must be adjusted depending on the process time and microlens materials.
Table I compares characteristics of the fabricated PDMS microlenses and photoresist molds
cured at two different conditions of room temperature for 24 hours and 85 °C for 1 hour. As
we can see in Table I, no shape change was observed between the two curing conditions.
When the photoresist mold was treated at 85 °C solely without PDMS casting, it became
severely distorted from its original lens shape. On the other hand, with PDMS, the shapes of
the photoresist mold and PDMS microlens were not deformed at all in spite of the high curing
temperature even after 7 times of PDMS replication at 85 °C. We speculated that the highly
viscous liquid PDMS (viscosity: 3100 cP) suppressed the reflow of the photoresist mold
effectively. Also, the narrow inlet of the photoresist mold cured at 85 °C was not deformed
during the 7 times of PDMS peel-off process as shown in Table I. This is believed to be due to
the elastic property of the PDMS. AFM measurement results showed that the surface
roughness of the photoresist mold after PDMS peel-off was improved from 4.8 nm to 2.6 nm
by elevating the curing temperature. The PDMS microlens is expected to have almost the
same surface roughness as the photoresist mold since the PDMS replicates the photoresist
mold surface roughness at the sub-nanometer scale [14]. As a result, the total integrated
(C) 2004 OSA 13 December 2004 / Vol. 12, No. 25 / OPTICS EXPRESS 6370
#5799 - $15.00 US Received 23 November 2004; revised 4 December 2004; accepted 6 December 2004
6. scattering (TIS) of the fabricated microlens surface is expected to be below 1% for visible
light [15], which is small enough to be used in most optical applications. The various
polymers which need a high curing temperature can be used as microlens materials as well to
achieve satisfactory microlens characteristics in varying applications. Only the PDMS
shrinkage (about 2.2 %) due to its relatively large thermal coefficient of expansion (310
ppm/°C)2
needs to be considered at the microlens design step to obtain a precise microlens
array.
Finally, the array of the circular gold pattern with a diameter of 20 µm on a glass was
observed through the fabricated microlens array with a width and a space of 50 µm. Figure 5
shows the microscope photographs of the gold patterns seen through the fabricated microlens
array, clearly showing magnification done by the microlens array.
Fig. 5. Optical photograph of the magnified gold patterns on a glass through the fabricated microlens
array.
4. Summary
To summarize, a simple plastic microlens array fabrication method was proposed and
investigated. In order to form various 3D photoresist molds for the microlens array, a 3D
diffuser lithography technology was developed, based on conventional lithography technology
with novel usage of a diffuser. The proposed method has three major advantages over
conventional methods. First, the process is very simple. Second, the microlens shape can be
controlled precisely and reproducibly by changing the diffuser and lithography process
parameters. Finally, the microlens array with a fill-factor of almost 100% can be fabricated in
a lithographical way, which is difficult to achieve with conventional technologies. The
fabricated versatile microlens array can be used for diverse optical systems with maximized
efficiency due to its various and controllable shapes and high fill-factor.
Acknowledgments
This work was supported by Brain Korea 21 program of the Ministry of Science and
Technology in Korea. The authors would like to thank Prof. Y.-K. Choi and Prof. Y.-H. Lee at
KAIST for their helpful comments.
(C) 2004 OSA 13 December 2004 / Vol. 12, No. 25 / OPTICS EXPRESS 6371
#5799 - $15.00 US Received 23 November 2004; revised 4 December 2004; accepted 6 December 2004